Multifunctional Cross-Linked Polymeric Membranes for Safe, High

Subscriber access provided by UNIV OF NEW ENGLAND ARMIDALE

Article

Multifunctional cross-linked polymeric membranes for safe, high-performance lithium batteries Sanjuna Stalin, Snehashis Choudhury, Kaihang Zhang, and Lynden A. Archer Chem. Mater., Just Accepted Manuscript • DOI: 10.1021/acs.chemmater.7b05353 • Publication Date (Web): 20 Feb 2018 Downloaded from http://pubs.acs.org on February 20, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Chemistry of Materials is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Chemistry of Materials

Multifunctional cross-linked polymeric membranes for safe, high-performance lithium batteries Sanjuna Stalin, Snehashis Choudhury, Kaihang Zhang, Lynden A. Archer Department of Chemical and Biomolecular Engineering, Cornell University, Ithaca, New York 14853, USA

Abstract Lithium ion batteries (LIBs) are the dominant power source in portable device technology and are poised to play a similar, dominant role in electrified transportation systems. While significant strides have been made in recent years in evolving the cathode chemistry and electrolyte formulations to meet higher energy storage demands, almost all contemporary LIB designs rely on flammable electrolyte solvents that are fundamentally unsafe. Here we report a facile UV cross-linking chemistry that can be used to create ion-conducting polymer networks containing dangling chains that impart specific, desired functionalities to liquid electrolytes. We show in particular that incorporation of mono-functional sulfonate and phosphate species in a photo/heatinitiated cross-linking reaction of a multifunctional oligomer provides a straightforward route to mechanically robust membranes able to transform transport properties and flammability of standard liquid electrolytes incorporated in their pores. We evaluate the physical and mechanical properties of the materials and on that basis report that dangling chains in the membrane pores are able to transform electrolyte safety features, without compromising performance in electrochemical cells. Such cross-linked membranes with different pendant groups covalently tethered to an ion-conducting framework are argued to provide an important platform for enhancing lithium battery safety and performance.

1 ACS Paragon Plus Environment

Chemistry of Materials 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Introduction Since their introduction in the 1980s Lithium ion batteries (LIBs) have found widespread applications in portable devices and are now poised to play as pivotal a role in enabling electrified transportation technologies 1–4. Despite the relatively modest storage capacities intrinsic to the graphitic anode and intercalating cathode chemistries used in LIBs, researchers have over the last two decades made steady progress in increasing the overall energy density of LIBs to the point where electric vehicles powered by LIBs have emerged as competitive technology in markets throughout the world. These advances have been enabled through fundamental understanding of inorganic chemistry of the cathode, which has lead to steady increases in the operating potential of LIBs. It is remarkable that this progress has been possible in cell designs based on essentially the same volatile and combustible liquid electrolytes employed in the first-generation LIBs5–10. Accidents reported across the world in both hand-held devices and electric vehicle technology has brought new attention to the safety features of electrolytes in LIBs and underscore the need for fundamental materials chemistry approaches that address limitations of LIB electrolytes. It is known that when a LIB is abused by electrochemical, mechanical, or thermal means, an irreversible thermal runaway reaction occurs that leads to catastrophic cell failure, which may end in fire or explosion. This process is known to originate from the high volatility and poor thermal stability of liquid electrolyte solvents in current use and to be facilitated by degradation of the oxygen-rich cathodes, particularly chemistries that result in the highest cell potentials and capacities. Although the mechanisms through which an abused cell fails by thermal runaway reactions are poorly understood, it is known that vaporization of the solvent and disproportionation of oxygen-rich materials in the cathode to produce oxygen fuel these thermal


Page 2 of 23

Page 3 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


events, leading to runaway reactions involving both solution and vapor phase ingredients11,12. Thus, it is difficult to design additive packages able to arrest thermal run away. Additionally, LIB cathode chemistries that hold greatest promise for improving storage capacity are arguably the least safe because they participate as oxygen source and in some cases catalyst in the combustion process. There has been an upsurge in recent interest in batteries based on solid-state electrolytes that are considered inherently non-flammable. Wong.et.al , for example, reported that perfluoropolyether based electrolytes with a high Li+ transference number13 are simultaneously able to improve the oxidative stability of ether-based electrolytes at the high cathode potentials of interest in contemporary LIBs, and to maintain non flammability in abused cells. Other recent studies, including work by Agrawal et al. reported highly conductive, bidisperse nanoparticle hybrid electrolytes which could be formulated to exhibit low or no flammability with tunable mechanical properties14. These works are however silent about the challenges of maintaining high enough room temperature ionic conductivity in the electrolyte bulk and at the solid electrolyte interphases in the absence of a molecular solvent to facilitate ion transport. Numerous studies have considered design rules for liquid electrolytes that preserve the beneficial transport properties of currently used electrolyte systems, but which are at the same time nonflammable. In this regard, significant efforts have been directed at developing strategies for incorporating flame retardants in liquid electrolytes15–17. Flame retardants of broadest interest function by a free radical scavenging mechanism, which limits the number density in the vapor phase of free radical species resulting from oxidation of liquid electrolyte and thereby terminate propagation of combustion reactions induced by abuse of liquid electrolytes 18. Phosphate and halogen based compounds have emerged as particularly interesting materials because they have



been proven to be effective in suppressing flammability of many of the liquids used in LIBs. Phosphates are often preferred over halogen-based materials due to their environmental benignity19. However, utilization of phosphate-based flame-retardants as electrolyte additives comes with their own drawbacks including reduced bulk and interfacial ionic conductivity in liquid electrolytes and accelerated levels of capacity fading. Efforts have been concentrated in engineering the cell components in order to prevent interference of the flame retardant with the cell performance. Among these efforts, the work by Liu et.al. stands out for its demonstration of electrospun core-shell microfiber separators that encapsulate the retardant Triphenyl Phosphate, which is released when the shell layer melts in response to a thermal trigger20. Recently, Wang et.al reported electrolytes in which a common flame-retardant, Trimethyl Phosphate, is used as the electrolyte solvent and high concentration of LiFSA/NaFSA salt used to achieve high room temperature ionic conductivity in Lithium and Sodium ion batteries21. Here, motivated by a different design concept — namely that the large surface area available in the pores of a battery separator can be used to covalently tether desirable functional groups.22 By appropriately choosing these groups we show that it is possible to impart ion-rectification and improved safety features in liquid electrolytes. Specifically, we design and synthesize crosslinked membranes in which flame-retardant molecules are covalently linked to the membrane as dangling chains. Application of the membranes as separators in LIBs is hypothesized to produce a protective “jacket” of sorts that enhances overall safety of the electrolyte. Comparison of electrolytes based on our design with those in which a similar flame-retarding agent is incorporated into the electrolyte bulk by physical mixing show that covalent tethering offers multiple benefits, including stable cell cycling, long-term flame retardancy23, and more straightforward strategies for tuning overall safety features of the electrolyte. We also


Page 4 of 23

Page 5 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


investigate the effect of these pendant molecular functionalities in the cross-linked membranes on physical, chemical and thermal properties of liquid electrolytes imbibed in the membranes and show that such systems provide simple chemical routes for improving LIB electrolyte safety features without compromising other beneficial characteristics. Results and Discussions Scheme (1) illustrates the approach used for creating the multifunctional cross-linked membranes used in the study. Previously, we reported synthesis of cross-linked polymer membranes with pendant sulfonate ionomers and showed that when swollen with liquid electrolytes, the ionomers produce an electrostatic field throughout the electrolyte, allowing one to regulate transport and shuttling of polysulfide anions in Lithium-sulfur batteries 24. Scheme 1 shows how this approach can be extended to incorporate multiple immobilized functionalities in an electrochemical cell. By immobilizing desirable functional groups to a membrane (separator) it is possible to largely isolate these functional groups from electrochemical processes at the electrodes, maximizing their beneficial effects in the electrolyte. Cross-linking was achieved in solution by exposing mixtures of poly (ethylene glycol) dimethacrylate (PEGDMA), Vinylsulfonic acid sodium salt (VS), the flame retardant (Diethyl Ally Phosphate-DEAP) and the photo-initiator methyl benzoylformate (MBF) in DMSO to UV light. It is known that upon UV excitation MBF undergoes a fast bond cleavage reaction to generate free radicals able to initiate polymerization of molecules containing vinyl/acrylate groups. A unique property of PEGDMA is that upon free-radical initiation, it can make up to four covalent linkages with other molecules, which leads to spontaneous formation of crosslinked or branched molecules25. By performing the reaction in a mixture of difunctional poly (ethylene glycol) dimethacrylate (PEGDMA) and monofunctional vinylsulfonic acid sodium salt



(VS) monomers we previously showed that it is possible to create cross-linked PEO networks containing covalently linked pendant sulfonate ionomer groups24. By varying the relative fractions of the di- and mono-functional monomers, the composition of dangling groups and structure of the cross-linked networks formed by PEGDMA can be facilely adjusted to create materials with a broad range of physical and mechanical properties. Here, by integrating a third monomer DEAP we show that the concept can be extended further to create networks that host multiple pendant functionalities simultaneously.

Scheme 1. Schematic diagram depicting synthesis procedure for cross-linked PEGDMA membranes with pendant sulfonate and phosphate groups. The ionic cross-links between the sulfonate chains are also depicted

Successful cross-linking of the monofunctional and difunctional monomers was verified by XRay Photoelectron Spectroscopy (XPS). Prior to this and all analytical processes reported in the paper, the as prepared membranes were soaked, thoroughly rinsed in DI water, and dried. The


Page 6 of 23

Page 7 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


rinsing step is an important part of the protocol as it facilitates removal of un-reacted monomer, isolated PEG and phosphate polymers not covalently linked to the network, and salt generated in the synthesis. The XPS spectra of membranes with pendant functional groups exhibit P2p and S2p peaks as shown in Fig.1(a). As expected, these peaks are absent in the spectra of pure crosslinked PEGDMA membrane in Fig.1(b).

Figure 1. Experimental verification of cross-linking mechanism (a) XPS analysis spectrum with pendant functional groups,(b) without functional groups (c) Phosphorus 2p peaks from fitting of XPS spectrum, and (d) Sulfur 2p peaks with background (green) from Shirley fitting, SEM images of (e) Multifunctional cross-linked membrane, (f) Pure cross-linked PEGDMA membrane



The P2p3/2 (133.2 eV) and P2p1/2 (134.6 eV) peaks in Fig.1(c) can be attributed to the Phosphate groups in DEAP26 and S2p3/2 peaks at 167.2eV and 168.6 eV in Fig.1(d) are characteristic of SO32- groups27. These results are consistent with those from complementary FTIR spectroscopic analysis reported in Fig.(S1) and discussed in the Supporting Materials section. The morphology of the synthesized membranes as seen under the Scanning Electron Microscope (SEM) is illustrated in Fig. 1(e) and 1(f), respectively, for the multifunctional membrane and the cross-linked analog that does not contain dangling chains. Large rope-like, intertwined polymer networks are clearly evident from the SEM micrographs for the pure cross-linked PEGDMA materials. The average thickness of the structures is large (~10µm), implying that the rope-like networks are formed from many molecules assembled together into cross-links. An important benefit of this unusual morphological feature is that the materials are amorphous; no evidence of an amorphous to crystalline phase transition is observed in DSC analysis of the dry, bulk membranes (see Fig (S2)). Introduction of monofunctional VS and DEAP monomers to the network is seen to reduce both the thickness and length of the rope-like structures, instead forming finer interconnected structures. This is tentatively attributed to viscosity or polymerization induced aggregation and rearrangement of PEGDMA chains as a result of ionic linkages between the sulfonate ionomers28,29. The uniformity and amorphous nature of the membranes proves advantageous in wettability and mechanical properties, as seen later. The thermal properties of the multifunctional membranes and the retardant molecule DEAP were studied in detail using Thermogravimetric Analysis (TGA). Typical results are reported in Fig (S3). The residue weight percentage in case of the multifunctional membrane matches well with the ionomer content, which forms a Li2SO4 salt residue upon thermal degradation in air. The corresponding derivative plots also provide information about the various components


Page 8 of 23

Page 9 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


present on the membrane and their rate of thermal degradation compared to the monomers before cross-linking. For further studies, the previously optimized PEGDMA:VS monomer mole ratio of 10:4 was used throughout, and the DEAP content was varied to see the influence on conductivity and mechanical properties before analyzing the flame retardant properties. The DC ionic conductivity was determined as a function of temperature for various compositions of functional groups in the multifunctional membrane after soaking in 1M LiPF6-EC/DMC (1:1) electrolyte for 24 hours. The temperature-dependent conductivity is found to be well fitted with the Vogel–Fulcher–Tammann (VFT) equation, implying that there are no melting or crystallization transitions of the soaked membranes in the temperature range tested. This finding is consistent with the DSC thermograms reported in Fig.(S4) and confirms that the materials are amorphous. As seen from Fig 2(a), all membranes exhibit high conductivity values of the order ̴ 3mS/cm at room temperature and increasing the fraction of dangling chains results in increased ionic conductivity. We attribute this observation to the increased wettability of the membranes with pendant chains, whereby the nucleophilic phosphate groups and the ionic linkages between the SO3- ionomers increase permeability and dielectric constant of the membranes. This finding is in agreement with previously reported results, which show that pendant short branches with particular chemical properties when introduced in a cross-linked polymer network enhance CO2 permeability by increasing the local free volume available for gas dissolution and penetration30.



Figure 2. Effect of pendant functionalities on wettability and mechanical properties. (a) Arrhenius plot of multifunctional membranes of different compositions (b) Glass transition characteristics, and (c) Storage and Loss modulus as a function of frequency

Measurements of the glass transition temperature provide a more quantitative approach for evaluating this hypothesis. When a polymer is wetted by a smaller molecular species with lower Tg, the decrease in glass transition temperature of the polymer follows a well-established dependence on the degree of plasticization31,32. Fig 2(b) reports the absolute glass transition temperature of different membranes and corresponding ∆Tg, which is the decrease in glass transition temperature after soaking in the liquid electrolyte. As the fraction of pendant groups increase in the network, we see that the Tg decreases and ∆Tg increases, corresponding to higher electrolyte uptake. The soaked membranes also exhibited distinct melting and crystallization peaks (Fig. (S4)), due to the melting transition of EC in the electrolyte. It is noteworthy that the peak heights at the transitions increases with increasing fraction of pendant groups, consistent


Page 10 of 23

Page 11 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


with the proposal that the electrolyte uptake is increased by incorporation of larger fraction of pendant chains in the networks. A high storage modulus is an important requirement for a LIB separator. The storage (G’) and loss modulus (G’’) were determined from torsional shear analysis of the dry and solvent-swollen membranes at small strain amplitude ( γ = 0.1%) (Fig. 2 (c)). It is seen that in all cases the dynamic storage modulus G’ is more than an order of magnitude larger than the dynamic loss G” modulus and nearly independent of frequency, indicating that the materials are solid-like whether swollen or not swollen by solvent. It is also apparent that the storage modulus for the pure crosslinked PEGDMA membrane is high (60 MPa) and the introduction of pendant groups lower the modulus. This is consistent with our previous statement that pendant chains in the network increases available local free volume. However, we also believe that the resilience of the mechanical properties in networks made more defective by incorporation of pendant groups reflects a partial compensation of elasticity by ionic cross-links induced by the sulfonate ionomers. Soaking the membranes in liquid electrolytes results in an additional factor of 3 reduction in modulus, but even then the elastic modulus of the materials (5MPa ) is still highenough to function as a LIB separator. On that basis, we conclude that a mole ratio of 10:4:10 (PEGDMA:VS: DEAP) is optimal with respect to ionic conductivity and storage modulus. In order to understand how the pendant chains impact safety characteristics and electrochemical performance of batteries, we compared flame retardance and electrochemical properties of the electrolyte-swollen membranes with those of electrolytes in which the same flame retardant is used as an additive. There is already a large body of work on how flame retardant additives adversely impact electrochemical performance of LIBs 33,17,34,35. For convenience, we first studied DEAP as the flame retardant additive for 1M LiPF6 in EC/DMC. The flame retarding



mechanism is believed to be due to a free-radical scavenging reaction, whereby DEAP releases PO● and PO2● which capture H● and OH● radicals from burning of the electrolyte in the vapor phase and terminate the chain reaction that produces combustion36. The self-extinguishing time measured for electrolytes with different amounts of DEAP are reported in Fig 3(a). To perform these measurements, glass fiber mats were soaked in each electrolyte for a preset time and the mass change recorded. Following the soaking period, mats were removed and ignited using a butane torch held at a fixed distance from the mat. The time required for the initial flame to burn out divided by the mass of electrolyte taken-up by the mat defines the self-extinguishing time for the electrolyte. As expected, the time to self-extinguish decreases with increasing DEAP content and the electrolyte is considered non-flammable when 25 wt% or more DEAP is added. The Arrhenius plot for the various compositions is shown in Fig 3(b). Increase in DEAP content leads to decrease in conductivity, though the decrease is still within the acceptable range of conductivity values for Lithium ion battery electrolytes.


Page 12 of 23

Page 13 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3. Physical incorporation of flame retardants in liquid electrolytes. (a) Self Extinguishing time of electrolytes with different fractions of DEAP in 1M LiPF6 in EC/DMC (b) Arrhenius plot depicting decreasing DC ionic conductivity with increasing flame retardant concentration (c) Voltage Profiles of electrolyte with 25% DEAP in a LiFePO4/Graphite full cell and (d) Nyquist plots for different compositions of DEAP in electrolyte

Electrochemical analysis of the electrolytes were performed using LiFePO4/Graphite full cells. In these cells 100μL of 1M LiPF6 in EC/DMC (50/50 v/v) with 25 wt % of DEAP was used as electrolyte and commercial Celgard 3501 was used as the separator and the cells were subjected to galvanostatic cycling at a fixed rate of C/3. The voltage profiles reported in Fig 3(c) shows large changes from as early as the 2nd cycle, and continues degradation of storage capacity in subsequent cycles. A conventional explanation for this observation is that the flame-retardant additive leads to formation of mechanically poor interfaces at the LIB anode where DEAP is subject to continuous reduction. To confirm this, impedance measurements were performed in a symmetric Li/Li cell to study the interfacial transport behavior of DEAP near a highly reducing electrode. Fig 3(d) shows the Nyquist plots for various compositions of DEAP in the electrolyte. It is clear from these results that the interfacial resistance increases with increasing DEAP content. Having confirmed the disadvantages of improving safety by using the flame retardants as additives in a liquid electrolyte, we studied how immobilizing the flame retardant on the separator renders it non-flammable without compromise on the electrochemical performance. Fig.4(a) reports the electrochemical stability window for the soaked multifunctional membranes with different compositions obtained from linear sweep voltammetry. It is well known that PEO breaks down close to 4V37 and this is seen from the onset of oxidation at ̴ 4V for cross-linked PEGDMA. The oxidation stability is seen to increase, by at least 1 V (Stability Window of the electrolytes increases from approximately 4V to 5V) in case of the multifunctional membranes. We attribute this result to rectification of anion transport by the negative charges imparted by the



sulfonate ionomers to the membranes. Our previous results show that whether such charges are present in a bulk membrane in the inter-electrode space24 or as a nanometer thin interfacial phase coated on the separator38 or electrode,39-40 they are highly effective in rectifying transport of negatively charged species in liquid electrolytes. Here we believe that the charged membranes serve to simultaneously limit access of anions to the highly oxidizing electrode surface and to constrain oxidation products from the PEO degradation to a self-limiting region near the electrode, limiting wholesale, bulk decomposition of the PEO electrolyte.

Figure 4. Electrochemical characterization of multifunctional membranes (a) Cyclic Voltammetry results of soaked multifunctional membranes with Pure PEGDMA cross-linked membrane as reference with a scan rate of 10mV/s, (b)Nyquist plots for different membranes and commercial Celgard seperator soaked in electrolyte (c) Voltage profiles for different cycles in a LiFePO4/Graphite full cell with multifunctional membrane (PEGDMA:VS:DEAP-10:4:10) soaked in electrolyte (d) Discharge Capacity and Coulombic Efficiency of full cell with soaked multifunctional membrane and electrolyte with DEAP additive for 100 cycles

Impedance spectroscopy was utilized to study the interfacial behavior in a symmetric cell with lithium as the two non-blocking electrodes. The Nyquist plots for the multifunctional membranes are shown in Fig. 4(b). Results are consistent with DC ionic conductivity measurements,


Page 14 of 23

Page 15 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


whereby increasing the fraction of pendant groups results in reduced bulk as well as interfacial impedance. The inset figure shows impedance data for pure PEGDMA cross-linked membrane soaked in electrolyte and the high bulk and interfacial impedance is a result of the poor wettability and electrolyte uptake of the membrane without any pendant groups. When directly compared to the impedance of the electrolyte with DEAP as an additive, we see that the multifunctional membranes have a more favorable interface with lithium, due to the prevention of reduction of the fire retardant on lithium. To evaluate the application of the multifunctional membranes as electrolyte/separator in a lithium ion battery, LiFePO4-Graphite full cells were constructed with membranes soaked in 1M LiPF6 in EC/DMC electrolyte without any additives. Fig. 4(d) shows the cycling data at a rate of C/3 for the multifunctional membrane electrolyte as well as DEAP physically mixed in electrolyte for comparison. The multifunctional membrane proved stable without any additives in the full cell, cycling stably for 100 cycles with >99% coulombic efficiency. This is in stark contrast compared to the electrolyte with physically mixed DEAP, whereby, as seen earlier, the irreversible capacity loss from the 1st cycle is evident with only 25% capacity retention in the first charge-discharge cycle. This further proves the stability of the flame retardant when immobilized on the separator, whereby it is no longer able to participate in side reactions with the graphite anode and interfere with ion transport at the interface. The stable voltage profiles (Fig. 4(c)) show relatively low overpotential at a rate of C/3 for each cycle, and this can be attributed to good wettability of the membrane and it’s single-ion conducting nature. The high voltage stability of the multifunctional membrane electrolyte was evaluated with a commercial high voltage cathode, in this case, LCO. Full cells with LCO as the cathode and graphite as the anode were assembled with the multifunctional membrane and Celgard in the



control case for comparison, after soaking in electrolyte for 24 hours. Cycling data and voltage profiles(inset) are reported in Fig.(S5) for a cycling rate of C/3 over 50 cycles. Comparable performance to Celgard is achieved in case of multifunctional membrane. The overpotential is seen to be slightly higher compared to the control case and this may be due to the higher interfacial resistance in case of the membrane compared to Celgard as seen previously. This can be better controlled by reducing the thickness of the membrane to values comparable to the standard Celgard membrane used in Lithium-ion cells. Finally, the thermal stability of the membranes soaked in electrolyte was analyzed by measuring the self-extinguishing times using the same flame testing process detailed earlier. All separators were soaked in 1M LiPF6 EC/DMC electrolyte prior to test. As seen from Fig.5, Celgard with electrolyte combusts on ignition and is destroyed in merely few seconds. Though cross-linked PEGDMA proves to be less readily combustible compared to Celgard, it still catches fire due to the flammability of the electrolyte and its intrinsic tendency to combust at elevated temperatures. The multifunctional membrane on the other hand is observed to be non-flammable due to the large number of pendant phosphate chains incorporated in the membranes, which suppress electrolyte combustion. This concept can also be applied to create flame-retardant membranes based on commonly used commercial polypropylene separators by first coating it with the precursor material and initiating the cross-linking reaction. As shown in Fig. 5(d) separators infused with cross-linked membrane at the optimal composition are non-flammable.


Page 16 of 23

Page 17 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Fig 5. Flame retardant properties of soaked multifunctional membranes compared to commercial seperator (a) Celgard, (b) Pure cross-linked PEGDMA, (c) Multifunctional membrane and (d) Multifunctional membrane as a coating on commercial separator

Conclusions In conclusion, we report a facile route to prepare cross-linked multifunctional membranes with multiple pendant functionalities to improve safety and electrochemical performance of LIB electrolytes, simultaneously. The physical and mechanical properties of the membranes are studied and found to be easily tuned by changing the fraction of pendant chains employed in the synthesis. By means of self-extinguishing flame studies and electrochemical analysis, the



membranes are shown to be highly efficient in improving safety characteristics of LIB electrolytes, while at the same time overcoming the well-known problem of interference of fire retardant additives with cell performance. As a final demonstration, we show that the same chemistry employed to create optimal free-standing membranes can be used to incorporate flame retardancy to a conventional polyolefin separator by using it as a support for the cross-linking reaction of the membrane components. Materials and Methods Membrane preparation: Multifunctional membranes were prepared using a scheme similar to previously reported. PEGDMA(Mw=750) and Vinylsulfonic acid sodium salt solution (25 wt% in H2O) and Diethyl Ally phosphate (DEAP) were purchased from Sigma Aldrich. VS salt was dried in vacuum oven for 24 hours and dissolved in DMSO (25 wt%). PEGDMA and DEAP were added to the DMSO solution in the desired molar ratio and stirred till a uniform solution was obtained. 5% Methyl benzoylformate (Sigma Aldrich) photoinitiator was added to the solution and the solution was cast on a Teflon plate for cross-linking under UV light (VMR UVAC 115 V ∼60 Hz 254/365 nm) for 20 minutes. Membranes of 100µm were harvested and washed with DI water to remove the DMSO and unreacted monomers. The membranes were then kept in a 3M LiOH solution for ion-exchange between Na+ and Li+. After 24 hours the membranes were washed with DI water and kept under vacuum for 24 hours at 80oC to remove excess water. The membranes were then punched out and stored in an Argon-glovebox before use. Characterization: The membranes were characterized by X-ray Photoelectron Spectroscopy using Surface Science Instruments SSX-100 with operating pressure of ~2×10-9 torr.


Page 18 of 23

Page 19 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Monochromatic Al K-α x-rays (1486.6eV) with beam diameter of 1mm were used. Photoelectrons were collected at an emission angle of 55° and the electron kinetic energy was determined by a hemispherical analyzer, where a pass energy of 150V was used for wide survey scans and 50V for high resolution scans. CasaXPS software was used for XPS data analysis with Shelby backgrounds and the spectra were referenced to adventitious C 1s at 284.5 eV. Morphologies of the membranes were studied using a LEO 1550 FESEM (Keck SEM). Mechanical properties of the membranes were investigated by dynamic mechanical analysis (DMA) using a TA Instruments DMA Q800. Thermogravimetric analysis (TGA) was used to determine the thermal degradation of different compositions of the membranes. Impedance versus frequency was measured using a Novocontrol N40 broadband dielectric spectroscopy instrument. Ionic conductivity as a function of frequency was also measured with a Novocontrol N40 broadband spectrometer fitted with a Quarto temperature control system. The samples were swelled in the electrolyte for 24 hours in the glovebox. The swollen membranes were punched into ½ inch diameter discs and sandwiched between stainless steel plates for conductivity measurements. Self-extinguishing time measurements were done by igniting the membranes/separators of interest, after soaking in electrolyte, using a butane torch at a fixed distance. The time to self-extinguish was measured and normalized using the mass of the soaked material. Electrochemical Characterization: 2030 coin-type cells were assembled in a glovebox (MBraun Labmaster) using LiFePO4 (2mAh/cm2, MTI corp.) or LCO (NOHMS Technologies) as the cathode and Graphite (NOHMS Technologies) as the anode. The loading on the electrode materials were 2mAh/cm2. The prepared membranes swollen with 1M LiPF6 in EC/DMC(50/50 v/v) served as the separator/electrolyte and Celgard 3501 was used in the control case. The room



temperature cycling characteristics of the cells were evaluated under galvanostatic conditions using Neware CT-3008 battery testers, and electrochemical processes in the cells were studied by linear sweep voltammetry using a CHI600D potentiostat. Associated Content Supporting information Supporting information is available free of charge on the ACS Publications website. Supporting information contains results from FTIR, TGA, DSC, electrochemical cycling analysis of membranes with a range of sulfonate and DEAP contents. Acknowledgements This work was supported by the National Science Foundation, Division of Materials Research, through Award No. DMR–1609125.

References 1.

Armand, M. & Tarascon, J. M. Building better batteries. Nature 451, 652–657 (2008).

2.

Tarascon, J. M. & Armand, M. Issues and challenges facing rechargeable lithium batteries. Nature 414, 359–367 (2001).

3.

Goodenough, J. B. & Kim, Y. Challenges for rechargeable Li batteries. Chem. Mater 22, 587–603 (2010).

4.

Van Noorden, R. The rechargeable revolution: A better battery. Nature 507, 26–28 (2014).

5.

Wang, Q. et al. Thermal runaway caused fire and explosion of lithium ion battery. J. Power Sources 208, 210–224 (2012).

6.

Bandhauer, T. M., Garimella, S. & Fuller, T. F. A Critical Review of Thermal Issues in LithiumIon Batteries. J. Electrochem. Soc. 158, R1-R25 (2011).

7.

Chen, S. C., Wan, C. C. & Wang, Y. Y. Thermal analysis of lithium-ion batteries. J. Power Sources 140, 111–124 (2005).

8.

Balakrishnan, P. G., Ramesh, R. & Prem Kumar, T. Safety mechanisms in lithium-ion batteries. J. Power Sources 155, 401–414 (2006).

9.

MacNeil, D. D., Lu, Z., Chen, Z. & Dahn, J. R. A comparison of the electrode/electrolyte reaction at elevated temperatures for various Li-ion battery cathodes. J. Power Sources 108, 8–14 (2002).

10.

Hammami, A., Raymond, N. & Armand, M. Lithium-ion batteries: Runaway risk of forming toxic compounds. Nature 424, 635–636 (2003).

11.

Golubkov, A. W. et al. Thermal-runaway experiments on consumer Li-ion batteries with metaloxide and olivin-type cathodes. RSC Adv. 4, 3633–3642 (2014).


Page 20 of 23

Page 21 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


12.

Hausbrand, R. et al. Fundamental degradation mechanisms of layered oxide Li-ion battery cathode materials: Methodology, insights and novel approaches. Mater. Sci. Eng. B Solid-State Mater. Adv. Technol. 192, 3–25 (2015).

13.

Wong, D. H. C. et al. Nonflammable perfluoropolyether-based electrolytes for lithium batteries. Proc. Natl. Acad. Sci. 111, 3327–3331 (2014).

14.

Agrawal, A., Choudhury, S. & Archer, L. A. A highly conductive, non-flammable polymer– nanoparticle hybrid electrolyte. RSC Adv. 5, 20800–20809 (2015).

15.

Xu, K., Zhang, S., Allen, J. L. & Jow, T. R. Evaluation of Fluorinated Alkyl Phosphates as Flame Retardants in Electrolytes for Li-Ion Batteries: II. Performance in Cell. J. Electrochem. Soc. 150, A170-A175 (2003).

16.

Xu, K., Ding, M. S., Zhang, S., Allen, J. L. & Jow, T. R. An Attempt to Formulate Nonflammable Lithium Ion Electrolytes with Alkyl Phosphates and Phosphazenes. J. Electrochem. Soc. 149, A622-A626 (2002).

17.

Tsujikawa, T. et al. Characteristics of lithium-ion battery with non-flammable electrolyte. J. Power Sources 189, 429–434 (2009).

18.

Levchik, S. V. A Review of Recent Progress in Phosphorus-based Flame Retardants. J. Fire Sci. 24, 345–364 (2006).

19.

Rakotomalala, M., Wagner, S. & Döring, M. Recent developments in halogen free flame retardants for epoxy resins for electrical and electronic applications. Materials (Basel). 3, 4300– 4327 (2010).

20.

Liu, K. et al. Electrospun core-shell microfiber separator with thermal-triggered flame-retardant properties for lithium-ion batteries. Sci. Adv. 3, e1601978 (2017).

21.

Wang, J. et al. Fire-extinguishing organic electrolytes for safe batteries. Nat. Energy 1–8 (2017). doi:10.1038/s41560-017-0033-8

22.

Zhang, W. et al. Design principles of functional polymer separators for high-energy metal-based batteries. Small 1703001 (2017). https://doi.org/10.1002/smll.201703001

23.

Saucǎ, S., Giamberini, M. & Reina, J. A. Flame retardant phosphorous-containing polymers obtained by chemically modifying poly(vinyl alcohol). Polym. Degrad. Stab. 98, 453–463 (2013)

24.

Ma, L., Nath, P., Tu, Z., Tikekar, M. & Archer, L. A. Highly Conductive, Sulfonated, UV-CrossLinked Separators for Li-S Batteries. Chem. Mater. 28, 5147–5154 (2016).

25.

Bäckström, S. et al. Tailoring Properties of Biocompatible PEG-DMA Hydrogels with UV Light. Mater. Sci. Appl. 3, 425–431 (2012).

26.

Briggs, D. Handbook of X‐ray Photoelectron Spectroscopy CD Wanger, WM Riggs, LE Davis, JF Moulder and GE Muilenberg Perkin‐Elmer Corp., Physical Electronics Division,. Surface and Interface Analysis 3, v–v (1981).

27.

Baltrusaitis, J., Cwiertny, D. M. & Grassian, V. H. Adsorption of sulfur dioxide on hematite and goethite particle surfaces. Phys. Chem. Chem. Phys. 9, 5542 (2007).

28.

Schulze, M. W., McIntosh, L. D., Hillmyer, M. A. & Lodge, T. P. High-modulus, highconductivity nanostructured polymer electrolyte membranes via polymerization-induced phase separation. Nano Lett. 14, 1,122–126 (2014).



29. 30.

Porcarelli, L., Gerbaldi, C., Bella, F. & Nair, J. R. Super Soft All-Ethylene Oxide Polymer Electrolyte for Safe All-Solid Lithium Batteries. Sci. Rep. 6, (2016). Kusuma, V. A., Freeman, B. D., Borns, M. A. & Kalika, D. S. Influence of chemical structure of short chain pendant groups on gas transport properties of cross-linked poly(ethylene oxide) copolymers. J. Memb. Sci. 327, 195–207 (2009).

31.

He, R. & Kyu, T. Effect of Plasticization on Ionic Conductivity Enhancement in Relation to Glass Transition Temperature of Cross-linked Polymer Electrolyte Membranes. Macromolecules 49, 5637–5648 (2016).

32.

Shen, M. C. & Tobolsky, A. V. Glass transition temperature of polymers, in Plasticization and Plasticizer Processes Chapter 2, 27–34 (1965) doi:10.1021/ba-1965-0048.ch002.

33.

Hu, J. et al. A new phosphonamidate as flame retardant additive in electrolytes for lithium ion batteries. J. Power Sources 197, 297–300 (2012).

34.

Ciosek Högström, K. et al. Impact of the flame retardant additive triphenyl phosphate (TPP) on the performance of graphite/LiFePO4 cells in high power applications Dedication: Professor per Jacobsson in memoriam. J. Power Sources 256, 430–439 (2014).

35.

Wang, Q., Sun, J., Yao, X. & Chen, C. 4-Isopropyl Phenyl Diphenyl Phosphate as FlameRetardant Additive for Lithium-Ion Battery Electrolyte. Electrochem. Solid-State Lett. 8, A467 (2005).

36.

Scharte, B. Phosphorus-based flame retardancy mechanisms-old hat or a starting point for future development? Materials (Basel). 3, 4710–4745 (2010).

37.

Osada, I., De Vries, H., Scrosati, B. & Passerini, S. Ionic-Liquid-Based Polymer Electrolytes for Battery Applications. Angew. Chem. Int. Ed. 55, 2, 500–513 (2016).

38.

Kim, M.S., Kim, M-S., Do, V., Lim, Y-R., Nah, I. W., Archer, L. A., Cho, W. I. Designing solid-electrolyte interphases for lithium sulfur electrodes using ionic shields. Nano Energy 41, 573-582 (2017).

39.

Lu, Y., Tikekar, M. D., Mohanty, R., Hendrickson, K. & Archer, L. A. Stable cycling of lithium metal batteries using high transference number electrolytes. Adv. Energy Mater. 5, 1402073 (2015).

40.

Tu, Z., Choudhury, S., Zachman, M. J., Wei, S., Zhang, K., Kourkoutis, L. F. & Archer, L. A. Designing artificial solid-electrolyte interphases for single-ion and high efficiency transport in batteries. Joule 1, 1-13 (2017).


Page 22 of 23

Page 23 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


TOC graphic:


Multifunctional Cross-Linked Polymeric Membranes for Safe, High

Recommend Documents